Add support for local large language models, (Ollama)

[iplayfast]

I have a network computer running linux, that I've set up an https://ollama.ai service on.

The steps to start the ollama service are:

installing & running ollama which is a llm server
installing litellm which is an openai emulator
running it.

litellm -m ollama/mistral --port 9000 --drop_params&

Which means I now have an openai compatible port at 192.168.0.15:9000

Unfortunately when I tried using the ChatGPT wizard, it came back with an error "No answer, try again"
On the linux side the message said:
Docs: https://docs.litellm.ai/docs/proxy_server

Worker Initialized

INFO: Application startup complete.
INFO: Uvicorn running on http://0.0.0.0:9000 (Press CTRL+C to quit)
192.168.0.196:50629 - "POST / HTTP/1.1" 405 Method Not Allowed

google show others with this problem.
openai/chatgpt-retrieval-plugin#74

[AliDehbansiahkarbon]

Hi dear Chris,
First of all, I appreciate you sharing the information with me, thank you.

This is not an issue man 🫡, so I moved it into the discussions tab for now.
For your information, I have been exploring several open-source projects like this over the past three months.
I am waiting to find a powerful/suitable local project to integrate with this plugin.
This one looks really cool, and I'm going to start working on it,
So stay tuned🫵 something greater will...